Cartograflow, for make flowmap

Françoise Bahoken

2019-03-17

Introduction

Cartograflow is designed to easily create thematic origin-destination (OD) flowmaps, based on the articulation of a matrice (an OD flow dataset) and a background map (a spatial shape).

This Vignette presents different functions that are mainly used to prepare the flowdata set. The spatial objects processing are those of {sp} and the mapping elements are often those of {Cartography} except, for example, the design of arrows.

Two matrice formats are available

– List format : “L” for “list” is a .csv 3 column flow dataset (origin, destination, flow_value) ;

– Matrice format : “M” for “Matrice” is a .csv [n*n] flow dataset.

– Use flowtabmat() in order to transform a “L” to “M” matrice format and vice versa.

In the case you only have a list of spatial units (code,X,Y), you can also generate an empty square matrice with flowtabmat()

At least two datasets are required

– a statistical dataset : a .csv “L” or “M” flow dataset ;

– a geographical dataset :

1. Preparing flow dataset

1-1. Pre-processing

– flowcarre() is to transform an un-square to a square matrice from a list of spatial objets ID (code)

– flowjointure() is to performs a spatial join between a flow dataset and a map background

– flowtabmat() is to transform a “M” matrice format to a “L” format and vice versa

– flowstructmat() fixes an unpreviously ID shift in the flow dataset “M” format. If necessary this function is to be used with flowjointure() and flowtabmat.

1-2. Computing flows

It is to decide firstly to zero or not the diagonal, see {base::diag}.

– flowtype() is to compute the main types of flows from an asymmetric flow dataset (matrice or list format). The result is a bilateral gross or net flows matrice.

It is also possible to compute the matrice’s margins in order to calculate probabilities of sending and receiving flows or all kinds of indicators. Use for that the R {base} or {dplyr}

2. Flow analysis

2.1. Concentration

– flowgini() performs a concentration analysis of a flow dataset - To be use before flowanalysis()

– flowanalysis() is to be used after flowgini() for computing a flow filter based on a double criterion for selecting flows before mapping :

or

2.2. Distance travelled

You have two ways to consider the distance travelled by flows : – if you have a matrice distance, go directly to flowreduct() at §2.2.3 ;

– if not, you can continue here, and have to choose the type of metric (continous or ordinal)

2.2.1. Compute continuous distances matrices

– flowjointure() performs an attribute spatial join - by origin (i) and by destination (j) - between a flow dataset and a spatial shape in order to transfert the origin-destination coordinates (Xi, Yi, Xj, Yj) of the base map to the flow matrice.

– flowdist() Computes a continous distance matrice choosing metric (only euclidian for this first version)

Computes on the resulting continous distance matrice at least a {base::summary} in order to select a distance filter criterion, then a flowreduct() and finally the flowmap() function.

2.2.2. Compute ordinal distances matrices

– flowcontig() compute an ordinal distance distance matrice based on a k-contiguity matrice. (k) is the order parameter, it is the number of borders to be crossed between origins and destinations places.

  • ordre=1, origin-destination places are adjacent ;

  • ordre=2, origin-destination places are distant from 2 borders ;

  • ordre=4, origin-destination places are distant from 4 borders.

Use after flowreduct() function and directly flowmap() without applying the filter parameter.

You can also map the neighbourhood spatial graph using flowmap() without applying the filter parameter.

2.2.3. Reducting a flow matrice by an external matrice

– flowreduct() is to perform the reduction of the flow dataset according to another matrice (especially a matrice distance)

3. Flow mapping

flowmap() is to create a layer of lines and plot them, using a flow dataset and a spatial shape.

3.1. Selecting what to map (if necessary)

– filtre is to choose to map all or selected matrice’s cases.

3.2. Filtering flow values

– seuil is to apply a numerical global (across the whole matrice) flow selection criterion.

You can apply an empirical criterion or use a robust one in the statistical meaning.

For robust filter criterion, use :

3.3. Filtering flow features

3.3. Making a flowmap collection

For doing a map collection, one can use for example :

3.4. Setting up the flow features

– taille is to modify the width of the origin-destination lines.

– a.fleche is to determine the kind of arrows to be drawn from places (in, out, in and out).

– a.length is to parameter the length of the arrow head in relation to the body of the arrow.

– a.angle is to parameter an angle from the shaft of the arrow to the edge of the arrow head.

– a.col is to parameter the color of the arrows.

– a.alpha is to parameter of opacity

3.4. Flowmap cartographic template

To dress and decorate the map by addin a legend layer, a north or south arrow, a scale, etc., see :

– {cartography::layoutLayer}

– {graphics::legend}

4. Example of applications

Inter-municipality professional mobility in Greater Paris.

4.1 Load packages

rm(list=ls())

library(dplyr)
library(sf)
library(cartography)
library(cartograflow)

4.2 Data

– Statistical dataset : extraction from the “Mobilités professionnelles en 2015 : déplacements domicile - lieu de travail” from the french census (Recensement de la population) - Base flux de mobilité, file.

-URL : https://www.insee.fr/fr/statistiques/fichier/3566008/rp2015_mobpro_txt.zip

– Geographical dataset :

4.2.1 Load Statistical information

data<-read.csv2("./data/tab_MGP_mobpro.csv",header=TRUE,sep=";",stringsAsFactors=FALSE,
                encoding="UTF-8",dec=".",check.names=FALSE)
head(data)
##   CODGEO                   LIBGEO  DCLT                    L_DCLT
## 1  75101 Paris 1er Arrondissement 75102  Paris 2e  Arrondissement
## 2  75101 Paris 1er Arrondissement 75105  Paris 5e  Arrondissement
## 3  75101 Paris 1er Arrondissement 75108  Paris 8e  Arrondissement
## 4  75101 Paris 1er Arrondissement 75109  Paris 9e  Arrondissement
## 5  75101 Paris 1er Arrondissement 75112 Paris 12e  Arrondissement
## 6  75101 Paris 1er Arrondissement 75113 Paris 13e  Arrondissement
##   NBFLUX_C14_ACTOCC15P
## 1                  247
## 2                  104
## 3                  426
## 4                  263
## 5                  123
## 6                  139
tabflow<-data%>% 
            select(CODGEO,DCLT,NBFLUX_C14_ACTOCC15P)
colnames(tabflow)<-c("i","j","Fij")

head(tabflow)
##       i     j Fij
## 1 75101 75102 247
## 2 75101 75105 104
## 3 75101 75108 426
## 4 75101 75109 263
## 5 75101 75112 123
## 6 75101 75113 139
str(tabflow)
## 'data.frame':    4692 obs. of  3 variables:
##  $ i  : int  75101 75101 75101 75101 75101 75101 75101 75101 75101 75101 ...
##  $ j  : int  75102 75105 75108 75109 75112 75113 75115 75116 92012 92026 ...
##  $ Fij: int  247 104 426 263 123 139 134 123 128 139 ...
#typage des variables
tabflow$i<-as.character(tabflow$i)
tabflow$j<-as.character(tabflow$j)
tabflow$Fij<-as.numeric(tabflow$Fij)

ID_CODE<-read.csv2("./data/COD_GEO_MGP.csv",header=TRUE,sep=";",stringsAsFactors=FALSE,encoding="UTF-8",dec=".",check.names=FALSE)
head(ID_CODE)
##   COD_GEO
## 1   75101
## 2   75102
## 3   75103
## 4   75104
## 5   75105
## 6   75106
CODE<-ID_CODE%>%select(COD_GEO)
colnames(CODE)<-c("CODGEO")

#CODE = as.data.table(CODE)
head(CODE)
##   CODGEO
## 1  75101
## 2  75102
## 3  75103
## 4  75104
## 5  75105
## 6  75106
#str(CODE)

#typage des variables
#CODE$CODGEO<-as.character(CODE$CODGEO)

4.2.2 Pre-processing flow dataset

#change matrice format (if necessary)
matflow <-flowtabmat(tabflow,matlist="M")
## Using Fij as value column: use value.var to override.
## warning:your matrix is not square!
head(matflow[1:4,1:4])
##       75101 75102 75103 75104
## 75101     0   247     0     0
## 75102   286     0   115     0
## 75103   379   410     0   224
## 75104   232   207   183     0
dim(matflow)
## [1] 141 183
#The Warning says that the matrice is note square.Dimension is 141x183 that is why we make it square, using cartograflow::flowcarre
matflow.sq<-flowcarre(CODE,tabflow,"i","j","Fij",format = "M",diagonale = TRUE,empty.sq = FALSE)
## 'data.frame':    150 obs. of  1 variable:
##  $ CODGEO: int  75101 75102 75103 75104 75105 75106 75107 75108 75109 75110 ...
## Using ydata as value column: use value.var to override.
## great:your matrix is square!
head(matflow.sq[1:4,1:4])
##       75101 75102 75103 75104
## 75101     0   247     0     0
## 75102   286     0   115     0
## 75103   379   410     0   224
## 75104   232   207   183     0
dim(matflow.sq)
## [1] 150 150

4.2.3 Computing flow dataset

# zero the diagonal of matrice format (if necessary)
# mat_nodiag<-matflow.sq

# diag(mat_nodiag) <- 0
# head(mat_nodiag[1:4,1:4])

# re change matrice format
# tabflow_nodiag <-flowtabmat(tab=mat_nodiag,
#                   matlist="L")
# colnames(tabflow_nodiag) <- c("i","j","Fij")
# head(tabflow_nodiag)

4.2.4 Computing main types of flow dataset

#matflow <-mat_nodiag

# Compute bilateral volume - from a "M" format
  matflow_vol<-flowtype(tab=matflow.sq,
                        format="M","bivolum")

# Compute bilateral volume - from a "L" format
  tabflow.sq<-flowtabmat(matflow.sq,matlist = "L")
  
# Compute only bilateral volume or net flows in one 3 column "L" format matrice
  # FSij will be the gross Fij flow values
  
  tabflow_vol<-flowtype(tabflow,
                        format="L","bivolum")
  head(tabflow_vol)
##       i     j FSij
## 1 75101 75102  533
## 2 75101 75105  460
## 3 75101 75108  727
## 4 75101 75109  742
## 5 75101 75112 1003
## 6 75101 75113 1110
  # FDij will be the net Fij flow values
  tabflow_sold<-flowtype(tabflow,
                        format="L","bisold")
  head(tabflow_sold)
##       i     j FDij
## 1 75101 75102   39
## 2 75101 75105  252
## 3 75101 75108 -125
## 4 75101 75109  216
## 5 75101 75112  757
## 6 75101 75113  832
#compute all types in one 6 columns "L" format matrice
tabflow_all<-flowtype(tabflow, 
                      format="L", 
                      x="all")
head(tabflow_all) 
##       i     j Fij Fji FSij FDij
## 1 75101 75102 247 286  533   39
## 2 75101 75105 104 356  460  252
## 3 75101 75108 426 301  727 -125
## 4 75101 75109 263 479  742  216
## 5 75101 75112 123 880 1003  757
## 6 75101 75113 139 971 1110  832
#compute and flow asymetry
tabflow_all$FAsy<-(tabflow_all$FDij / tabflow_all$FDij)*100

4.3 Flow mapping

Data= “tabflow.sq” Basemap= “MGP_communes.shp”

4.3.1 Direct flow mapping

knitr::opts_chunk$set(fig.width=6, fig.height=6)

# Plot all theoretical OD links (no filter)
flowmap(tab = tabflow.sq,
               format="L",
               fdc="./data/MGP_communes.shp",
               code="IDCOM",
               filtre=FALSE)
## All theorical links are plotted
mtext("All theoretical relations",side = 3)

#Plot all existing relations
flowmap(tab = tabflow.sq,
                 format="L",
                 fdc="./data/MGP_communes.shp",
                 code="IDCOM",
                 filtre=TRUE,
                 a.col="#3f4247",
                 seuil=1, #flow values up to 1
                 taille=1,
                 a.fleche = 0)

mtext("All existing relations ",side = 3)

#Plot only flow values up to 100 (the french INSEE's)
flowmap(tab = tabflow.sq,
                 format="L",
                 fdc="./data/MGP_communes.shp",
                 code="IDCOM",
                 filtre=TRUE,
                 a.col="#3f4247",
                 seuil=100,   #flow value > 100
                 taille=5,     #graphic parameteter
                 a.fleche = 1, #add arrow
                 a.length = 0.11)

mtext("Existing flows up to 100",side = 3)

4.5 Mapping a robust flow selection

4.5.0 Statistical parameter

tabflow.sq<-tabflow.sq %>% filter(tabflow.sq$ydata!=0)

#extraction des quantiles
summary(tabflow.sq$ydata)
##    Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 
##   100.0   145.0   228.0   357.1   403.0  4857.0
decile<-c(0,0.1,0.2,0.3,0.4,0.5,0.6,0.7,0.8,0.9,1)
quantile(tabflow.sq$ydata,decile)
##     0%    10%    20%    30%    40%    50%    60%    70%    80%    90% 
##  100.0  116.0  133.4  157.0  187.0  228.0  277.0  350.0  472.2  748.0 
##   100% 
## 4857.0
#Plot all flow up to Q3

par(mar=c(0,0,1,0))
extent <- c(2800000, 1340000, 6400000, 4800000)
resolution<-300

flowmap(tab = tabflow.sq,
                 format="L",
                 fdc="./data/MGP_communes.shp",
                 code="IDCOM",
                 filtre=TRUE,
                 a.col="#3f4247",
                 seuil=748,   #flow value > 748 (90% information)
                 taille=5,     #graphic parameteter
                 a.fleche = 1, #add arrow
                 a.length = 0.11)

mtext("Existing flows up to 748 - 90% information",side = 3)

4.5.1 Concentration analysis

# LIST FORMAT
#--------------------

tabflow.sq<-flowtabmat(matflow.sq,matlist = "L")

# 1- Computes Gini coefficient
#--------------------
tab_gini<-flowgini(tabflow.sq,
                   format="L",
                   origin="i",
                   dest="j",
                   valflow="ydata",
                   fdc = "./data/MGP_communes.shp",
                   code="IDCOM",
                   lorenz.plot = FALSE)
## coefficient de gini = 44.13 %
head(tab_gini)
##          i     j ydata       X1      Y1       X2      Y2 link     flowcum
## 1067 75117 75108  4857 649165.4 6865479 649575.7 6863852    1 0.003117114
## 1065 75115 75108  4852 648097.2 6860237 649575.7 6863852    1 0.006231019
## 1066 75116 75108  4000 645848.7 6862519 649575.7 6863852    1 0.008798130
## 1068 75118 75108  3560 652206.2 6866034 649575.7 6863852    1 0.011082859
## 2281 92012 75116  3263 644144.9 6859859 645848.7 6862519    1 0.013176979
## 4515 75115 92012  3142 648097.2 6860237 644144.9 6859859    1 0.015193445
##           linkcum
## 1067 0.0002292001
## 1065 0.0004584002
## 1066 0.0006876003
## 1068 0.0009168004
## 2281 0.0011460005
## 4515 0.0013752006
# 2- Plot Lorenz curve
#--------------------

flowgini(tab_gini, 
         format="L",
         origin="i",
         dest="j",
         valflow="ydata",
         fdc = "./data/MGP_communes.shp",
         code="IDCOM",
         lorenz.plot = TRUE)
# 3- Compute critflow parameter and flowmap
#-------------------------------
#critflow = 0.02 (2% of the largest flows)
flowanalysis(tab_gini,
             critflow = 0.02,
             result = "signif")
## [1] "seuil =  3070  ---  flows =  2 % ---  links =  0.18 %"
#seuil = 3070 - 2% flows - 0,1% links
par(mar=c(0,0,1,0))
flowmap(tabflow.sq,
        format="L",
        fdc="./data/MGP_communes.shp",
        code="IDCOM",
        filtre=TRUE,
        seuil=1176,
        taille=5,
        a.fleche = 1,
        a.length = 0.11,
        a.angle = 30,
        a.col="#3f4247")

mtext("significative flowmap : values up to 3070 - 2% flow information - 0,1% links",side = 3)

# 4- Compute critlink parameter and flowmap 
#-------------------------------

flowanalysis(tab_gini,
             critlink = 0.01,
             result = "density")
## [1] "seuil =  2045  ---  flows =  7.21 % ---  links =  1 %"
# Plot 4 % of the total features, flow greater than 2015
par(mar=c(0,0,1,0))
flowmap(tab = tabflow.sq,
        format="L",
        fdc="./data/MGP_communes.shp",
        code="IDCOM",
        filtre=TRUE,
        a.col="#3f4247",
        seuil=2045,
        taille=5,
        a.fleche = 1,
        a.length = 0.11,
        a.angle = 30
        )

mtext("Small density flowmap : values up to 2045 - 7,21 flow information  - 1%  links",side = 3)

4.5.2 Concentration analysis

library(cartography)

par(mar=c(0,0,1,0))
extent <- c(2800000, 1340000, 6400000, 4800000)
resolution<-300

# Flow map
flowmap(tabflow.sq,
        format="L",
        fdc="./data/MGP_communes.shp",
        code="IDCOM",
        filtre=TRUE,
        seuil=2045, # Min  value
        taille=4, #taille max du plus gros flux  
        a.fleche = 1,
        a.length = 0.11,
        a.angle = 30,
        a.col="#3f4247")

# Legend
##---------------
legendPropLines(pos="topleft",
                title.txt="Number of commuters",
                title.cex=1,    # taille du titre de la legende
                cex=0.8,
                values.cex= 0.7,  # size of the values in the legend.
                var=c(2045,max(tabflow.sq$ydata)), # # Min  - Max flow value
                lwd=4, #taille max du plus gros flux 
                frame = FALSE,
                col="#3f4247",
                values.rnd = 0
                )

layoutLayer(title = "Professional mobility in Greater Paris : the strongest 1%",
            author = "Cartograflow, 2019",
            sources = "Sources : data : INSEE, RP, MOBPRO, 2017 ; basemap : IGN - GEOFLA 2015, UMS 2414 RIATE, 2018",
            scale = 5,
            tabtitle = TRUE,
            frame = TRUE,
            north(pos = "topright"),
            col = "grey",
            coltitle ="black")

# Add territorial boundaries
fdc_ter <- st_read("./data/MGP_territoires.shp", stringsAsFactors = F)
## Reading layer `MGP_territoires' from data source `D:\R\package\cartograflow\vignettes\data\MGP_territoires.shp' using driver `ESRI Shapefile'
## Simple feature collection with 12 features and 14 fields
## geometry type:  POLYGON
## dimension:      XY
## bbox:           xmin: 2.145699 ymin: 48.64615 xmax: 2.615706 ymax: 49.01233
## epsg (SRID):    4326
## proj4string:    +proj=longlat +datum=WGS84 +no_defs
# Harmonizing projections
prj <- 2154
fdc_ter <- st_transform(fdc_ter, prj)

plot(fdc_ter$geometry, col=NA, border="#332d2e",lwd=1.9, add=T)

4.6 Thresholding flows by distance travelled

4.6.1 Continuous distance

tab<-flowjointure(tabflow.sq,"./data/MGP_communes.shp","IDCOM")
tab.distance<-flowdist(tab,dist.method = "euclidian",result = "dist")

#summarise (for example) to find mean, dmin or dmax 
summary(tab.distance$distance)
##    Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 
##       0    8879   13642   14135   18971   36867
#reduce the flow dataset from a selected distance travelled < 4506 m (corresponding to Q1)
tab.flow<-flowreduct(tabflow.sq,
                     tab.distance,
                     metric = "continous",
                     select = "dmax", #max distance parameter 
                     d = 4056)  #max distance value

#select for all i,j flow values up to 0
  flow.d<-tab.flow%>%
        select(i,j,flowfilter)%>%
        filter(flowfilter !=0)

flowmap(flow.d,format="L","./data/MGP_communes.shp","IDCOM",
        filtre = TRUE,
        a.col="#3f4247",
        a.length = 0.11,
        a.fleche =1)
## warning : you use the default threshold= 1
mtext("Professional mobility less than 4506 meters (Q1)")

## Final flowmap less than 4,5 km 

#dev.off() #nettoyage fenetre graphique

par(mar=c(0,0,1,0))

extent <- c(2800000, 1340000, 6400000, 4800000)
resolution<-300

flowmap(flow.d,format="L","./data/MGP_communes.shp","IDCOM",
        filtre = TRUE,
        taille = 4,
        a.col="#3f4247",
        a.length = 0.11,
        a.fleche =1)
## warning : you use the default threshold= 1
plot(fdc_ter$geometry, col=NA, border="black",lwd=1.9, add=T)

legendPropLines(pos="topleft",
                title.txt="Number of commuters",
                title.cex=1,    # taille du titre de la legende
                cex=0.8,
                values.cex= 0.7,  # size of the values in the legend.
                var=c(min(flow.d$flowfilter),max(flow.d$flowfilter)), # Min - Max de la serie
                col="#3f4247",
                lwd=4, #taille max du plus gros flux défini ds flowmap
                frame = FALSE,
                values.rnd = 0
                )
# Habillage

layoutLayer(title = "Professional mobility in Greater Paris : distance travelled less than 4.5 km",
            author = "Cartograflow, 2019",
            sources = "Sources : data : INSEE, RP, MOBPRO, 2017 ; basemap : IGN - GEOFLA 2015, UMS 2414 RIATE, 2018",
            scale = 5,
            tabtitle = TRUE,
            frame = TRUE,
            north(pos = "topright"),
            col = "grey",
            coltitle ="black")

4.6.2 Ordinal distance

## Neighbouring graph (ordre 1)
graph_ckij_1<-flowcontig("./data/MGP_communes.shp","IDCOM",ordre =1)

flowmap(graph_ckij_1,
        format="L",
        "./data/MGP_communes.shp","IDCOM",
        filtre = TRUE, 
        taille = 0.5)
## warning : you use the default threshold= 1
mtext("Neighbouring graph (ordre 1)",
      side=3)

## Reducing flow matrice by the neighbouring graph (ordre 1)
reduc<-flowreduct(tabflow.sq,graph_ckij_1,metric = "ordinal")
flow.c<-reduc %>%
  select(i,j,flux)%>%
  filter(flux!=0)

## Plot adjacent flows 
flowmap(flow.c,format="L","./data/MGP_communes.shp","IDCOM",
        filtre = TRUE,
        taille = 2,
        a.col="#3f4247",
        a.length = 0.11,
        a.fleche =1)
## warning : you use the default threshold= 1
mtext("Adjacent flows",
      side=3)

## Final adjacent flowmap 

#dev.off() #nettoyage fenetre graphique

par(mar=c(0,0,1,0))
extent <- c(2800000, 1340000, 6400000, 4800000)
resolution<-300

flowmap(flow.c,format="L","./data/MGP_communes.shp","IDCOM",
        filtre = TRUE,
        taille = 4,
        a.col="#3f4247",
        a.length = 0.1,
        a.fleche =1)
## warning : you use the default threshold= 1
# Add territorial boundaries
fdc_ter <- st_read("./data/MGP_territoires.shp", stringsAsFactors = F)
## Reading layer `MGP_territoires' from data source `D:\R\package\cartograflow\vignettes\data\MGP_territoires.shp' using driver `ESRI Shapefile'
## Simple feature collection with 12 features and 14 fields
## geometry type:  POLYGON
## dimension:      XY
## bbox:           xmin: 2.145699 ymin: 48.64615 xmax: 2.615706 ymax: 49.01233
## epsg (SRID):    4326
## proj4string:    +proj=longlat +datum=WGS84 +no_defs
prj <- 2154
fdc_ter <- st_transform(fdc_ter, prj)

plot(fdc_ter$geometry, 
     col=NA, 
     border="#332d2e",
     lwd=1.9, 
     add=T)

# Add Legend
legendPropLines(pos="topleft",
                title.txt="Number of commuters",
                title.cex=1,    # taille du titre de la legende
                cex=0.8,
                values.cex= 0.7,  # size of the values in the legend.
                var=c(min(flow.c$flux),max(flow.c$flux)), # Min - Max de la serie
                col="#3f4247",
                lwd=4, #taille max du plus gros flux défini ds flowmap
                frame = FALSE,
                values.rnd = 0
                )
# Habillage

layoutLayer(title = "Professional mobility in Greater Paris between neighbouring municipalities",
            author = "Cartograflow, 2019",
            sources = "Sources : data : INSEE, RP, MOBPRO, 2017 ; basemap : IGN - GEOFLA 2015, UMS 2414 RIATE, 2018",
            scale = 5,
            tabtitle = TRUE,
            frame = TRUE,
            north(pos = "topright"),
            col = "grey",
            coltitle ="black")

#mtext("Note : Only flows between adjacent municipalities are plot",
 #     side=2)

## Neighbouring graph (ordre 2)
graph_ckij_2<-flowcontig("./data/MGP_communes.shp","IDCOM",ordre =2)
head(graph_ckij_2)
##       i     j ydata
## 1 75102 75109     2
## 2 75103 75109     2
## 3 75104 75109     2
## 4 75105 75109     2
## 5 75106 75109     2
## 6 75107 75109     2
flowmap(graph_ckij_2,
        format="L",
        "./data/MGP_communes.shp","IDCOM",
        filtre = TRUE, 
        taille = 0.5)
## warning : you use the default threshold= 1
mtext("Neighbouring graph (order 2)",
      side=3)

## Reducing flow matrice by the neighbouring graph (ordre 2)
reduc2<-flowreduct(tabflow.sq,graph_ckij_2,metric = "ordinal")

flow.c2<-reduc2 %>%
         select(i,j,flux)%>%
         filter(flux!=0)

## Plot adjacent flows 
flowmap(flow.c2,format="L","./data/MGP_communes.shp","IDCOM",
        filtre = TRUE,
        taille = 4,
        a.length = 0.11,
        a.fleche =1)
## warning : you use the default threshold= 1
mtext("Flows between municipalities distant from 2 boundaries",
      side=3)

Final adjacent flowmap

library(sf)

## Final adjacent flowmap 
#dev.off() #nettoyage fenetre graphique

par(mar=c(0,0,1,0))

extent <- c(2800000, 1340000, 6400000, 4800000)
resolution<-300

flowmap(flow.c2,format="L","./data/MGP_communes.shp","IDCOM",
        filtre = TRUE,
        taille = 4,
        a.length = 0.11,
        a.fleche =1,
        a.col = "#3f4247")
## warning : you use the default threshold= 1
legendPropLines(pos="topleft",
                title.txt="Number of commuters",
                title.cex=1,    # taille du titre de la legende
                cex=0.8,
                values.cex=0.7,  # size of the values in the legend.
                var=c(min(flow.c2$flux),max(flow.c2$flux)), # Min - Max de la serie
                col="#3f4247",
                lwd=4, #taille max du plus gros flux défini ds flowmap
                frame = FALSE,
                values.rnd = 0
                )
#Habillage
layoutLayer(title = "Professional mobility in Greater Paris between municipalities from two boundaries",
            author = "Cartograflow, 2019",
            sources = "Sources : data : INSEE, RP, MOBPRO, 2017 ; basemap : IGN - GEOFLA 2015, UMS 2414 RIATE, 2018",
            scale = 5,
            tabtitle = TRUE,
            frame = TRUE,
            north(pos = "topright"),
            col = "grey",
            coltitle ="black")

#mtext("Note : Only flows between municipalities distant by 2 boundaries are plot",
#      side=2)

# Import de fonds
fdc_ter <- st_read("./data/MGP_territoires.shp", stringsAsFactors = F)
## Reading layer `MGP_territoires' from data source `D:\R\package\cartograflow\vignettes\data\MGP_territoires.shp' using driver `ESRI Shapefile'
## Simple feature collection with 12 features and 14 fields
## geometry type:  POLYGON
## dimension:      XY
## bbox:           xmin: 2.145699 ymin: 48.64615 xmax: 2.615706 ymax: 49.01233
## epsg (SRID):    4326
## proj4string:    +proj=longlat +datum=WGS84 +no_defs
prj <- 2154

fdc_ter <- st_transform(fdc_ter, prj)

# Add territorial limits

#plot(fdc_com$geometry, col=NA, border="blue",lwd=0.2)
plot(fdc_ter$geometry, col=NA, border="#332d2e",lwd=1.9, add=T)
#plot(fdc_lim$geometry,col=NA,border="red",lwd=1.5,add=T)

5.Reproducibility

sessionInfo()
## R version 3.5.2 (2018-12-20)
## Platform: x86_64-w64-mingw32/x64 (64-bit)
## Running under: Windows 10 x64 (build 15063)
## 
## Matrix products: default
## 
## locale:
## [1] LC_COLLATE=C                   LC_CTYPE=French_France.1252   
## [3] LC_MONETARY=French_France.1252 LC_NUMERIC=C                  
## [5] LC_TIME=French_France.1252    
## 
## attached base packages:
## [1] stats     graphics  grDevices utils     datasets  methods   base     
## 
## other attached packages:
## [1] bindrcpp_0.2.2          cartography_2.1.2       sf_0.7-2               
## [4] dplyr_0.7.8             cartograflow_0.0.0.9000
## 
## loaded via a namespace (and not attached):
##  [1] httr_1.4.0        pkgload_1.0.2     tidyr_0.8.2      
##  [4] jsonlite_1.6      viridisLite_0.3.0 shiny_1.2.0      
##  [7] assertthat_0.2.0  sp_1.3-1          yaml_2.2.0       
## [10] remotes_2.0.2     sessioninfo_1.1.1 pillar_1.3.1     
## [13] backports_1.1.3   lattice_0.20-38   glue_1.3.0       
## [16] digest_0.6.18     promises_1.0.1    colorspace_1.3-2 
## [19] httpuv_1.4.5.1    htmltools_0.3.6   plyr_1.8.4       
## [22] pkgconfig_2.0.2   devtools_2.0.1    xtable_1.8-3     
## [25] purrr_0.2.5       scales_1.0.0      processx_3.2.1   
## [28] later_0.7.5       tibble_2.0.0      ggplot2_3.1.0    
## [31] usethis_1.4.0     withr_2.1.2       lazyeval_0.2.1   
## [34] cli_1.0.1         mime_0.6          magrittr_1.5     
## [37] crayon_1.3.4      memoise_1.1.0     maptools_0.9-4   
## [40] evaluate_0.12     ps_1.3.0          fs_1.2.6         
## [43] xml2_1.2.0        foreign_0.8-71    class_7.3-14     
## [46] pkgbuild_1.0.2    tools_3.5.2       data.table_1.12.0
## [49] prettyunits_1.0.2 stringr_1.3.1     plotly_4.8.0     
## [52] munsell_0.5.0     callr_3.1.1       compiler_3.5.2   
## [55] e1071_1.7-0       rlang_0.3.0.1     classInt_0.3-1   
## [58] units_0.6-2       grid_3.5.2        rstudioapi_0.8   
## [61] htmlwidgets_1.3   crosstalk_1.0.0   labeling_0.3     
## [64] rmarkdown_1.11    testthat_2.0.1    gtable_0.2.0     
## [67] DBI_1.0.0         roxygen2_6.1.1    g.data_2.4       
## [70] reshape2_1.4.3    R6_2.3.0          knitr_1.21       
## [73] rgdal_1.3-6       rgeos_0.4-2       bindr_0.1.1      
## [76] commonmark_1.7    rprojroot_1.3-2   desc_1.2.0       
## [79] stringi_1.2.4     Rcpp_1.0.0        tidyselect_0.2.5 
## [82] xfun_0.4